In recent years, with development of digital transmission technology, digital broadcasting using satellites, cables, or terrestrial waves has been widely put into actual use. Particularly, the OFDM transmission system has already been put into practical use as digital terrestrial television broadcasting scheme in Europe, and it has also been decided in Japan that the OFDM method as digital terrestrial television broadcasting scheme and digital terrestrial audio broadcasting scheme would be adopted.
The OFDM transmission method assigns data to plural different carriers that are orthogonal to each other within a transmission band to perform modulation or demodulation. In this method, Inverse Fast Fourier Transform (hereinafter, referred to as IFFT) processing is performed on a transmitting end, and Fast Fourier Transform (hereinafter, referred to as FFT) processing is performed on a receiving end. Each carrier can employ an arbitrary modulation scheme, for example, synchronous modulation such as QPSK (Quaternary Phase Shift Keying) and QAM (Quadrature Amplitude Modulation), or differential modulation such as DQPSK (Differential Quaternary Phase Shift Keying).
In the synchronous modulation scheme, a pilot signal whose amplitude and phase are known on the receiving end is periodically inserted into a transmission signal and, on the receiving end, the channel characteristic is obtained with reference to the pilot signal, thereby performing demodulation. In the differential modulation scheme, demodulation is performed by delay detection. Further, in digital transmission methods such as the OFMD transmission method, error correction code decoding processing is carried out to improve the transmission characteristics.
However, when a transmission line includes frequency selectivity interference, whose major examples are multi-path interference that lowers the levels of specific carriers due to reflected waves, spurious interference that is produced for example in a case where a receiver mounted on a mobile unit is moving, and co-channel interference resulting from analog broadcasting that coexists with digital broadcasting, demodulation performance or error correction ability is greatly deteriorated.
An OFDM receiver that can avoid such situation has been already developed and disclosed in Patent Document 1 that will be described later. This prior art is briefly described with reference to drawings.
Initially, a construction of the prior art OFDM receiver is shown in FIG. 24. In this OFDM receiver, an OFDM transmission signal is inputted to a tuner unit 103 through a receiving antenna 101 and an RF amplifier 102, and station selection processing is carried out in the tuner unit. The station selection processing performed in the tuner unit 103 is implemented by tuning an oscillation frequency of a local oscillator 111 for a desired channel frequency in accordance with a frequency control signal that is inputted to a station selection information input terminal 110.
The output signal from the tuner unit 103 is converted into a digital signal by an analog-to-digital (hereinafter, referred to as A/D) conversion unit 104, and then subjected to orthogonal detection by an orthogonal detection unit 105, thereby obtaining a baseband OFDM signal. This baseband OFDM signal is supplied to a FFT unit 106. The FFT unit 106 transforms the inputted OFDM signal of time domain into a frequency domain signal. Here, an A/D conversion clock, and clocks or timing signals that are employed in other digital circuits are clocks or signals which are reproduced from the baseband OFDM signal by a synchronous reproduction unit 112.
The output signal from the FFT unit 106 indicates the phase and amplitude of the OFDM signal at each carrier, and this is inputted to a demodulation unit 107. The demodulation unit 107 demodulates the inputted OFDM signal by synchronous detection corresponding to the modulation scheme. By the synchronous detection, the channel characteristic of each carrier is detected using the pilot signal that is inserted at a rate of ⅓ in the frequency direction and ¼ in the time direction, and amplitude equalization and phase equalization are performed in accordance with the detected channel characteristics.
In the synchronous detection, pilot signals are arranged in 4 symbol periods in the received OFDM signal, and the channel characteristics of 3-carrier intervals are obtained from the pilot signal of 4-symbol period. Then, the channel characteristics of all carriers are obtained by interpolating these characteristics in the frequency direction. The demodulated signal is inputted to the error correction unit 108, errors that have occurred during the transmission are corrected, and the corrected signal is outputted from an output terminal 109.
The output signal of the FFT unit 106 is also inputted to the interference detection unit 113. The interference detection unit 113 judges the state of the received pilot signal to judge carriers which are subjected to frequency selectivity interference, and the judged result is inputted to the demodulation unit 107, the error correction unit 108, and the synchronous reproduction unit 112, whereby the demodulation performance is improved.
That is, because the demodulation unit 107 detects the channel characteristic for each carrier at the time of synchronous detection using the pilot signals, thereby performing the amplitude equalization and phase equalization thereof, when it is judged from the interference carrier information that the frequency that is subjected to the interference is equal to the frequency of the pilot signal, the demodulation unit does not use this signal, but detects a channel characteristic using another signal that has been interpolated with a pilot signal that is not subjected to the interference to perform demodulation. Further, the error correction unit 108 performs weighting processing such as loss correction with information of carriers that are subjected to interference. The synchronous reproduction unit 112 performs synchronous reproduction with a small error in accordance with a signal that is not subjected to interference.
FIG. 25 is a block diagram illustrating a specific construction of the interference detection unit 113 in the multi-carrier receiver of FIG. 24. A signal that has been subjected to fast Fourier transform is inputted from the FFT unit 106 to a pilot extraction unit 113a in the interference detection unit 113. The pilot extraction unit 113a extracts the pilot signal from the inputted signal. The output of the pilot extraction unit is inputted to an integrator 113b and a subtractor 113c. 
The integrator 113b integrates the amplitudes of respective pilot signals to obtain an average value, and outputs the average amplitude to the subtractor 113c. The subtractor 113c obtains a difference between the average amplitude of the pilot signals and the amplitude of each pilot signal. The obtained difference is inputted to an absolute value operation unit 113d as an error for each pilot signal. The absolute value operation unit 113d calculates the absolute value of the error for each pilot signal.
The output of the absolute value operation unit 113d is supplied to an integrator 113e, and integration processing for the error of each pilot signal in the time direction is executed. The result of this processing is supplied to a comparator 113f and an averaging unit 113g as an error signal for each pilot signal. Here, the error signal for each pilot signal corresponds to a C/N ratio for each pilot signal. The C/N ratio for each pilot signal is outputted from the averaging unit 113g as a C/N ratio for all pilot signals. On the other hand, the comparator 113f performs a comparison between the C/N ratio of each pilot signal and the C/N ratio of all pilot signals, and when a difference in the C/N ratio is large on the basis of the comparison result, it determines that there exists frequency selectivity interference. The output from the comparator 113f is inputted to the demodulation unit 107, the error correction unit 108, and the synchronous reproduction unit 112 as the above-mentioned interference carrier information.
The error correction unit 108 performs the weighting processing such as loss correction, in accordance with the information of carriers that are subjected to interference, thereby reducing the influence of the interference.
As described above, the conventional OFDM receiver monitors the pilot carriers of the received OFDM signal, and determines carriers subjected to the interference, thereby improving the demodulation performance.
When the OFDM signal of the digital terrestrial broadcasting is received, the frequency selectivity interference such as spurious interference or co-channel interference associated with analog television, or jumping-in of a clock of the receiver itself may occur. When the carriers are subjected to such interference, the performance of the demodulation error correction is substantially deteriorated.
In the prior art, it is conceivable that the output from the integrator 113e indicates the level of interference, while the level of interference becomes higher when the influence of the interference upon the OFDM signal is larger. Therefore, when frequency selectivity interference is included in the received OFDM signal, the output of the integrator 113e exhibits a prominent level near a position where the interference exists on the frequency axis.
FIG. 12 shows a state where two kinds of frequency selectivity interference are simultaneously superimposed upon the received OFDM signal. These two kinds of frequency selective interference are referred to as interferences A and B, respectively, and it is assumed that the interferences A and B are in proximity to different positions fA and fB on the frequency axis, respectively. It is further assumed that these interferences A and B are produced by different factors, respectively.
Then, a description will be given of how the conventional detection of interference changes according to a difference between the influences of the interferences A and B upon the received signal. FIGS. 26(a) and 26(b) show the output levels of the integrator 113e, and the average values outputted from the averaging unit 113g of FIG. 25, respectively. Initially, the description is given of a case where there is not much difference between the influences of the interference A and B upon the OFDM signal, as shown in FIG. 26(a). In this case, there is some difference between the output level of the integrator 113e and the average value in proximity to the positions fA and fB on the frequency axis, respectively. Therefore, on the basis of these differences, the comparator 113f can easily detect both of the interferences A and B.
Next, a description is given of a case where the influence of the interference A upon the OFDM signal is quite larger than the influence of the interference B upon the OFDM signal, as shown in FIG. 26(b). In this case, there is a sufficient difference between the output level of the integrator 113e and the average value near the position fA on the frequency axis, while, as compared to this, the difference between the output level of the integrator 113e and the average value near the position fB on the frequency axis is relatively smaller.
This is because the averaging unit 113g calculates the average value of all carriers (all pilot signals) regardless of the magnitude of the signal level (interference level) outputted from the integrator 113e, and accordingly when a signal whose level is locally prominent is outputted from the integrator 113e, the average value outputted from the averaging unit 113g becomes large under the influence of the locally high level.
Therefore, in the case as shown in FIG. 26(b), the comparator 113f easily detects the interference Abut has a great difficulty in detecting the interference B. Consequently, an opportunity that an improvement means which performs weighting or loss correction to carriers near the position fB, which are subjected to the interference B, achieves some effects is lost, whereby the modulation error correction performance may be deteriorated.
As described above, the prior art receiver judges the presence or absence of frequency selective interference on the basis of the average value of detected interference levels. Thus, when a plurality of frequency selective interference having difference influences are included, there is a possibility that interference having a relatively smaller influence may be undetected. Consequently, error correction is performed using carriers that are subjected to the interference, whereby demodulation error correction performance is deteriorated.
As another prior art, there is an interference detection correction method that is described in Patent Document 2 (hereinafter, referred to as a second prior art). In this second prior art, a distribution value (C/N ratio) of carriers is detected as an interference level.
In this second prior art, an average value of the distribution values in the frequency direction is calculated, and carriers having distribution values that are larger than the calculated average value are detected as carriers that are subjected to frequency selective interference, thereby performing loss correction. Also in this case, the presence or absence of the frequency selective interference is judged on the basis of the average value of detected interference levels as in the above-mentioned case, so that when there are a plurality of frequency selective interference having different degrees of influences, interference having a relative smaller degree of influence may not be detected. Accordingly, error correction is performed unfavorably using the carriers that to the interference, whereby the demodulation error correction performance may be deteriorated.
The second prior art further describes another method by which the minimum value of distribution values in the frequency direction is obtained, and loss correction is performed to carriers having larger values than this minimum value, as another interference detection/correction method. However, in this interference detection/correction method, a following erroneous operation may occur. That is, when a signal that includes no operation frequency selective interference and has been passed through a low C/N transmission line is received, the distribution values of the carriers may vary. In such case, when the minimum value of the distribution values in the frequency direction is obtained to perform loss correction for carriers that have larger values than the minimum value, carriers having larger distribution values than the minimum value are subjected to the loss correction. Consequently, there may be some carriers that are not affected by frequency selective interference but are subjected to loss correction processing for a reason that is different from the primary object. This may contrarily result in reduction of the modulation error correction performance.
The second prior art further describes, as still another interference detection/correction method, a method by which the average value and the minimum value of distribution values in the frequency direction are obtained, then a threshold level is set between the minimum value and the average value, and loss correction is performed to carriers having larger values than the threshold level. Also in this case, when there is frequency selective interference having a prominent interference level, the average value is unfavorably increased under the influence of this high level, and thus the threshold level may vary according to the states of interference. Consequently, when a plurality of frequency selective interference having different degrees of influences are included, interference having a relative smaller influence may not be detected, whereby the demodulation error correction performance may be deteriorated.
The second prior art further describes another method by which interference detection is performed utilizing a fact that the spectrum is known, or the interference level has a prominent value in the case of co-channel interference in the analog television broadcasting, as still another interference detection/correction method. However, even when this method is employed, appropriate interference detection algorithm cannot always be performed to frequency selective interference other than the co-channel interference in the analog television broadcasting.
As described above, even when the interference detection/correction method according to the second prior art is employed, there are some situations where appropriate interference detection and error correction cannot be carried out, depending on the interference state of the transmission line, resulting in degradation in demodulation or error correction ability.
Patent Document 1
                Japanese Published Patent Application No. Hei. 11-252040 (pages 4-5, FIG. 1)Patent Document 2        Japanese Patent No. 2954570 (pages 8-9, FIG. 2)        
The present invention is made in view of the above-mentioned conventional problems. This invention has for its object to provide a receiving apparatus and a receiving method that, when receiving a multi-carrier signal such as an OFDM signal and performing demodulation and error correction, can detect influences on the multi-carrier signal with great accuracy, and suppress deterioration in demodulation error correction performance even when there are a plurality of frequency selective interference having different degrees of influences.